Value Stream Map
Where does time die when someone turns a skill into a business?
Current State (Manual)
| Step | Activity | Time | Waste Type | Who Does It |
|---|---|---|---|---|
| 1 | Research domain, competitors, market size | 40+ hrs | Overproduction — every founder researches independently | Founder |
| 2 | Draft business plan, value proposition | 20+ hrs | Over-processing — custom document for commodity questions | Founder |
| 3 | Identify target customers (ICP) | 15+ hrs | Motion — switching between planning and selling contexts | Founder |
| 4 | Choose tech stack, set up infrastructure | 10+ hrs | Waiting — blocked on decisions that are commodity | Founder + Dev |
| 5 | Set up payments, legal, insurance | 20+ hrs | Transportation — knowledge doesn't transfer between founders | Founder |
| 6 | Design marketing, content strategy | 15+ hrs | Overproduction — same marketing playbook reinvented | Founder |
| 7 | Build sales pipeline, outreach templates | 10+ hrs | Inventory — shelved leads from untested channels | Founder |
| 8 | Set up operations, admin, accounting | 20+ hrs | Over-processing — custom ops for standard workflows | Founder |
| Total | Manual business creation | ~150+ hrs | 3-4 weeks minimum |
Failure rate: ~90% of startups fail. Most from operational infrastructure issues, not domain expertise. The person brilliant at their craft drowns in the 95% that has nothing to do with their skill.
Future State (Generator)
| Step | Activity | Time | Algorithm | Output |
|---|---|---|---|---|
| 1 | Input skill + domain + location | 5 min | — | venture_business_ideas row |
| 2 | AI domain research + competitor analysis | 15 min | eval-runner (quality gate) | Market intelligence report |
| 3 | Value proposition generation + scoring | 10 min | eval-runner, spcl-scoring | Scored value prop |
| 4 | ICP identification from research | 10 min | explore-exploit | Ranked customer segments |
| 5 | Go/no-go gate | 5 min | vfl-force-potential, optimal-stopping | Pass/fail with reasoning |
| 6 | Sprout Generator creates infrastructure | 4 min | — | Running venture app |
| 7 | Operations seeding (WorkCharts + sales + marketing + accounting) | 20 min | spcl-scoring | Operational business |
| Total | Generator business creation | ~1 hour | 6 algorithms | Operating venture |
The 95/5 Split
| Provider | What They Handle | % of Work |
|---|---|---|
| Mycelium | Market research, business model, operations templates, sales pipeline, marketing cadence, agent workforce, platform infrastructure | 95% |
| Founder | Their craft, their location, their instinct for what's missing | 5% |
Seven Wastes
| Waste | Current | Generator Eliminates |
|---|---|---|
| Overproduction | Every founder researches same competitors | AI research shared, algorithms learn from aggregate |
| Waiting | Blocked on commodity technical decisions | Platform decisions are defaults, not choices |
| Transportation | Knowledge dies with each failed venture | WorkChart receipts feed next venture's templates |
| Over-processing | Custom CRM, payments, marketing per venture | Shared platform, scoped by ventureId |
| Inventory | Shelved business plans that never executed | Optimal-stopping gate prevents premature building |
| Motion | Context-switching between planning and building | Single flow: input → score → build → operate |
| Defects | Bad market fit discovered after infrastructure investment | VFL scoring catches misfit before Sprout triggers |
Key Finding
150+ hours manual → ~1 hour generated (150x reduction). But the real value isn't speed — it's waste elimination. Each of the seven wastes compounds across every venture on the network. The generator doesn't just make one business faster — it makes the platform smarter with each venture that runs.
Context
- Outcome Map — Previous: what does success look like?
- Dependency Map — Next: what must happen first?
- Business Development — The framework mapped here